Dockery/Trilium
1
0
Fork 0
mirror of https://github.com/zadam/trilium.git synced 2026-05-07 07:26:36 +02:00
Code Issues Packages Projects Releases Activity

fix(ocr): adapt OfficeProcessor to officeparser v6.1.0 ESM changes

Browse Source 
v6.1.0 added native ESM with Node16 resolution and a strict exports
field, breaking deep subpath imports like
officeparser/dist/parsers/ExcelParser.js. Switch to the main package
entry and use parseOfficeAsync(), which accepts a Buffer and
auto-detects the format via magic bytes.

Co-authored-by: Elian Doran <eliandoran@users.noreply.github.com>
This commit is contained in:
claude[bot]
2026-04-18 17:07:02 +00:00
parent 189867ca03
commit b2bcccb4c7
1 changed files with 16 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

38
apps/server/src/services/ocr/processors/office_processor.ts
View File

@@ -1,25 +1,20 @@
import { parseExcel } from 'officeparser/dist/parsers/ExcelParser.js';
import { parseOpenOffice } from 'officeparser/dist/parsers/OpenOfficeParser.js';
import { parsePowerPoint } from 'officeparser/dist/parsers/PowerPointParser.js';
import { parseWord } from 'officeparser/dist/parsers/WordParser.js';
import type { OfficeParserConfig } from 'officeparser/dist/types.js';
import officeparser from 'officeparser';
import type { OfficeParserConfig } from 'officeparser';
import log from '../../log.js';
import { OCRProcessingOptions, OCRResult } from '../ocr_service.js';
import { FileProcessor } from './file_processor.js';
type Parser = (buffer: Buffer, config: OfficeParserConfig) => Promise<{ toText(): string }>;
const PARSER_BY_MIME: Record<string, Parser> = {
const SUPPORTED_MIME_TYPES = new Set([
// Office Open XML
'application/vnd.openxmlformats-officedocument.wordprocessingml.document': parseWord,
'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet': parseExcel,
'application/vnd.openxmlformats-officedocument.presentationml.presentation': parsePowerPoint,
'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
'application/vnd.openxmlformats-officedocument.presentationml.presentation',
// OpenDocument
'application/vnd.oasis.opendocument.text': parseOpenOffice,
'application/vnd.oasis.opendocument.spreadsheet': parseOpenOffice,
'application/vnd.oasis.opendocument.presentation': parseOpenOffice
};
'application/vnd.oasis.opendocument.text',
'application/vnd.oasis.opendocument.spreadsheet',
'application/vnd.oasis.opendocument.presentation'
]);
const PARSER_CONFIG: OfficeParserConfig = {
outputErrorToConsole: false,
@@ -30,29 +25,28 @@ const PARSER_CONFIG: OfficeParserConfig = {
/**
* Office document processor for extracting text from DOCX/XLSX/PPTX and ODT/ODS/ODP files.
* Uses individual parsers from officeparser v6 to avoid pulling in pdfjs-dist.
* Uses officeparser's main API, which auto-detects the format from the buffer's magic bytes.
*/
export class OfficeProcessor extends FileProcessor {
canProcess(mimeType: string): boolean {
return mimeType in PARSER_BY_MIME;
return SUPPORTED_MIME_TYPES.has(mimeType);
}
getSupportedMimeTypes(): string[] {
return Object.keys(PARSER_BY_MIME);
return [...SUPPORTED_MIME_TYPES];
}
async extractText(buffer: Buffer, options: OCRProcessingOptions = {}): Promise<OCRResult> {
const mimeType = options.mimeType;
if (!mimeType || !(mimeType in PARSER_BY_MIME)) {
if (!mimeType || !SUPPORTED_MIME_TYPES.has(mimeType)) {
throw new Error(`Unsupported MIME type for Office processor: ${mimeType}`);
}
log.info(`Starting Office document text extraction for ${mimeType}...`);
const parse = PARSER_BY_MIME[mimeType];
const ast = await parse(buffer, PARSER_CONFIG);
const trimmed = ast.toText().trim();
const text = await officeparser.parseOfficeAsync(buffer, PARSER_CONFIG);
const trimmed = text.trim();
return {
text: trimmed,